AutoML in Python¶

  • AutoViz
  • LazyPredict
  • Scikit-Optimize

Dataset loading¶

In [1]:
import os
os.getcwd()
Out[1]:
'D:\\PYTHON\\AutoML'
In [2]:
os.listdir()
Out[2]:
['.ipynb_checkpoints',
 'AutoML_practice.ipynb',
 'AutoML_practice2.ipynb',
 'AutoViz_Plots',
 'auto_ml.ipynb',
 'heart.csv',
 'plots_optimize',
 'README.md',
 'scikit_optimize.ipynb',
 'sklearn-gridsearchcv-replacement.ipynb',
 'test']
In [3]:
import pandas as pd

df = pd.read_csv("heart.csv")

df.head()
Out[3]:
age sex cp trestbps chol fbs restecg thalach exang oldpeak slope ca thal target
0 52 1 0 125 212 0 1 168 0 1.0 2 2 3 0
1 53 1 0 140 203 1 0 155 1 3.1 0 0 3 0
2 70 1 0 145 174 0 1 125 1 2.6 0 0 3 0
3 61 1 0 148 203 0 1 161 0 0.0 2 1 3 0
4 62 0 0 138 294 1 1 106 0 1.9 1 3 2 0
In [4]:
df.info()

<class 'pandas.core.frame.DataFrame'>
RangeIndex: 1025 entries, 0 to 1024
Data columns (total 14 columns):
 #   Column    Non-Null Count  Dtype  
---  ------    --------------  -----  
 0   age       1025 non-null   int64  
 1   sex       1025 non-null   int64  
 2   cp        1025 non-null   int64  
 3   trestbps  1025 non-null   int64  
 4   chol      1025 non-null   int64  
 5   fbs       1025 non-null   int64  
 6   restecg   1025 non-null   int64  
 7   thalach   1025 non-null   int64  
 8   exang     1025 non-null   int64  
 9   oldpeak   1025 non-null   float64
 10  slope     1025 non-null   int64  
 11  ca        1025 non-null   int64  
 12  thal      1025 non-null   int64  
 13  target    1025 non-null   int64  
dtypes: float64(1), int64(13)
memory usage: 112.2 KB
In [5]:
df.describe().transpose()
Out[5]:
count mean std min 25% 50% 75% max
age 1025.0 54.434146 9.072290 29.0 48.0 56.0 61.0 77.0
sex 1025.0 0.695610 0.460373 0.0 0.0 1.0 1.0 1.0
cp 1025.0 0.942439 1.029641 0.0 0.0 1.0 2.0 3.0
trestbps 1025.0 131.611707 17.516718 94.0 120.0 130.0 140.0 200.0
chol 1025.0 246.000000 51.592510 126.0 211.0 240.0 275.0 564.0
fbs 1025.0 0.149268 0.356527 0.0 0.0 0.0 0.0 1.0
restecg 1025.0 0.529756 0.527878 0.0 0.0 1.0 1.0 2.0
thalach 1025.0 149.114146 23.005724 71.0 132.0 152.0 166.0 202.0
exang 1025.0 0.336585 0.472772 0.0 0.0 0.0 1.0 1.0
oldpeak 1025.0 1.071512 1.175053 0.0 0.0 0.8 1.8 6.2
slope 1025.0 1.385366 0.617755 0.0 1.0 1.0 2.0 2.0
ca 1025.0 0.754146 1.030798 0.0 0.0 0.0 1.0 4.0
thal 1025.0 2.323902 0.620660 0.0 2.0 2.0 3.0 3.0
target 1025.0 0.513171 0.500070 0.0 0.0 1.0 1.0 1.0
In [6]:
df.nunique()
Out[6]:
age          41
sex           2
cp            4
trestbps     49
chol        152
fbs           2
restecg       3
thalach      91
exang         2
oldpeak      40
slope         3
ca            5
thal          4
target        2
dtype: int64
In [7]:
df.target
Out[7]:
0       0
1       0
2       0
3       0
4       0
       ..
1020    1
1021    0
1022    0
1023    1
1024    0
Name: target, Length: 1025, dtype: int64

Automatically Visualization¶

AutoViz 깃헙

https://github.com/AutoViML/AutoViz

usage

filename = ""
sep = ","
dft = AV.AutoViz(
    filename,
    sep=",",
    depVar="",
    dfte=None,
    header=0,
    verbose=0,
    lowess=False,
    chart_format="svg",
    max_rows_analyzed=150000,
    max_cols_analyzed=30,
    save_plot_dir=None
)

그려주는 그래프 종류¶

  • Bar_Plots: Top 20개의 값 sorting 하여 bar plot plotting(기준 column: oldpeak)
  • Bar_Plots_Pivots: 카테고리컬 데이터별 연속형 변수에 대한 값 조합
  • Box_Plots: oldpeak each target
  • Dist_Plots: 변수들의 분포도 Normed Histogram (density estimation plot)
  • Dist_Plots_target: target의 class 분포도
  • Heat_Maps: 각 변수들 간의 correlation plot
  • Scatter_Plots: target별 oldpeak 값의 scatter plot
  • Numeric Columns : ['oldpeak']
  • Integer-Categorical Columns: ['age', 'cp', 'trestbps', 'chol', 'restecg', 'thalach', 'slope', 'ca', 'thal']
  • String-Categorical Columns: []
  • Factor-Categorical Columns: []
  • String-Boolean Columns: []
  • Numeric-Boolean Columns: ['sex', 'fbs', 'exang']
  • Discrete String Columns: []
  • NLP text Columns: []
  • Date Time Columns: []
  • ID Columns: []

  • Categorical variables %s (" ['age', 'cp', 'trestbps', 'chol', 'restecg', 'thalach', 'slope', 'ca', " "'thal', 'sex', 'fbs', 'exang']")

  • Continuous variables %s " ['oldpeak']"

In [8]:
%matplotlib inline
In [9]:
from autoviz.AutoViz_Class import AutoViz_Class
import matplotlib.pyplot as plt
# plt.style.use("dark_background")

Imported v0.1.58. After importing, execute '%matplotlib inline' to display charts in Jupyter.
    AV = AutoViz_Class()
    dfte = AV.AutoViz(filename, sep=',', depVar='', dfte=None, header=0, verbose=1, lowess=False,
               chart_format='svg',max_rows_analyzed=150000,max_cols_analyzed=30, save_plot_dir=None)
Update: verbose=0 displays charts in your local Jupyter notebook.
        verbose=1 additionally provides EDA data cleaning suggestions. It also displays charts.
        verbose=2 does not display charts but saves them in AutoViz_Plots folder in local machine.
        chart_format='bokeh' displays charts in your local Jupyter notebook.
        chart_format='server' displays charts in your browser: one tab for each chart type
        chart_format='html' silently saves interactive HTML files in your local machine
In [10]:
AV = AutoViz_Class()

# 1. 기본 형태인 svg로 저장 
#    저장 폴더를 지정하지 않으면 자동으로 `AutoViz_Plots` 폴더를 생성하여 
#    출력물을 저장함
AV.AutoViz(filename = '',
          dfte = df,
          depVar = 'target',
          verbose = 2, # 0: 간단히 표시, 1: 자세히 표시, 2: 파일로 저장
          max_rows_analyzed = df.shape[0],
          max_cols_analyzed = df.shape[1])

Shape of your Data Set loaded: (1025, 14)
#######################################################################################
######################## C L A S S I F Y I N G  V A R I A B L E S  ####################
#######################################################################################
Classifying variables in data set...
Data cleaning improvement suggestions. Complete them before proceeding to ML modeling.
  Nuniques dtype Nulls Nullpercent NuniquePercent Value counts Min Data cleaning improvement suggestions
chol 152 int64 0 0.000000 14.829268 0
thalach 91 int64 0 0.000000 8.878049 0
trestbps 49 int64 0 0.000000 4.780488 0
age 41 int64 0 0.000000 4.000000 0
oldpeak 40 float64 0 0.000000 3.902439 0 skewed: cap or drop outliers
ca 5 int64 0 0.000000 0.487805 0
cp 4 int64 0 0.000000 0.390244 0
thal 4 int64 0 0.000000 0.390244 0
restecg 3 int64 0 0.000000 0.292683 0
slope 3 int64 0 0.000000 0.292683 0
sex 2 int64 0 0.000000 0.195122 0
fbs 2 int64 0 0.000000 0.195122 0
exang 2 int64 0 0.000000 0.195122 0 
  Printing upto 30 columns max in each category:
    Numeric Columns : ['oldpeak']
    Integer-Categorical Columns: ['age', 'cp', 'trestbps', 'chol', 'restecg', 'thalach', 'slope', 'ca', 'thal']
    String-Categorical Columns: []
    Factor-Categorical Columns: []
    String-Boolean Columns: []
    Numeric-Boolean Columns: ['sex', 'fbs', 'exang']
    Discrete String Columns: []
    NLP text Columns: []
    Date Time Columns: []
    ID Columns: []
    Columns that will not be considered in modeling: []
    13 Predictors classified...
        No variables removed since no ID or low-information variables found in data set
Since Number of Rows in data 1025 exceeds maximum, randomly sampling 1025 rows for EDA...

################ Binary_Classification problem #####################
   Columns to delete:
'   []'
   Boolean variables %s 
"   ['sex', 'fbs', 'exang']"
   Categorical variables %s 
("   ['age', 'cp', 'trestbps', 'chol', 'restecg', 'thalach', 'slope', 'ca', "
 "'thal', 'sex', 'fbs', 'exang']")
   Continuous variables %s 
"   ['oldpeak']"
   Discrete string variables %s 
'   []'
   Date and time variables %s 
'   []'
   ID variables %s 
'   []'
   Target variable %s 
'   target'
All Plots are saved in .\AutoViz_Plots\target
Time to run AutoViz = 7 seconds
Out[10]:
age sex cp trestbps chol fbs restecg thalach exang oldpeak slope ca thal target
807 44 1 2 130 233 0 1 179 1 0.4 2 0 2 0
27 58 0 1 136 319 1 0 152 0 0.0 2 2 2 1
77 63 1 0 140 187 0 0 144 1 4.0 2 2 3 1
406 58 1 2 140 211 1 0 165 0 0.0 2 0 2 0
886 61 1 0 120 260 0 1 140 1 3.6 1 1 3 1
... ... ... ... ... ... ... ... ... ... ... ... ... ... ...
835 49 1 2 118 149 0 0 126 0 0.8 2 3 2 1
192 67 0 2 115 564 0 0 160 0 1.6 1 0 3 0
629 65 1 3 138 282 1 0 174 0 1.4 1 1 2 1
559 67 1 0 120 237 0 1 71 0 1.0 1 0 2 1
684 60 1 2 140 185 0 0 155 0 3.0 1 0 2 1

1025 rows × 14 columns

In [11]:
# 2. png 타입의 파일로 저장
#    저장 위치도 지정 가능 

AV.AutoViz(filename = '',
          dfte = df, # 데이터 프레임을 입력 데이터로 사용 시 지정 
          depVar = 'target', # 타겟 변수 지정 
          verbose = 2, # 0: 간단히 표시, 1: 자세히 표시, 2: 파일로 저장
          max_rows_analyzed = df.shape[0], # 분석에 사용되는 최대 row 수 지정
          max_cols_analyzed = df.shape[1], # 분석에 사용되는 최대 column(feature) 수 지정
          chart_format = "png", # 생성(저장)되는 차트의 파일 포맷(확장자)
          save_plot_dir = "./plots_autoviz2_png") # 저장되는 폴더 지정 

Shape of your Data Set loaded: (1025, 14)
#######################################################################################
######################## C L A S S I F Y I N G  V A R I A B L E S  ####################
#######################################################################################
Classifying variables in data set...
Data cleaning improvement suggestions. Complete them before proceeding to ML modeling.
  Nuniques dtype Nulls Nullpercent NuniquePercent Value counts Min Data cleaning improvement suggestions
chol 152 int64 0 0.000000 14.829268 0
thalach 91 int64 0 0.000000 8.878049 0
trestbps 49 int64 0 0.000000 4.780488 0
age 41 int64 0 0.000000 4.000000 0
oldpeak 40 float64 0 0.000000 3.902439 0 skewed: cap or drop outliers
ca 5 int64 0 0.000000 0.487805 0
cp 4 int64 0 0.000000 0.390244 0
thal 4 int64 0 0.000000 0.390244 0
restecg 3 int64 0 0.000000 0.292683 0
slope 3 int64 0 0.000000 0.292683 0
sex 2 int64 0 0.000000 0.195122 0
fbs 2 int64 0 0.000000 0.195122 0
exang 2 int64 0 0.000000 0.195122 0 
  Printing upto 30 columns max in each category:
    Numeric Columns : ['oldpeak']
    Integer-Categorical Columns: ['age', 'cp', 'trestbps', 'chol', 'restecg', 'thalach', 'slope', 'ca', 'thal']
    String-Categorical Columns: []
    Factor-Categorical Columns: []
    String-Boolean Columns: []
    Numeric-Boolean Columns: ['sex', 'fbs', 'exang']
    Discrete String Columns: []
    NLP text Columns: []
    Date Time Columns: []
    ID Columns: []
    Columns that will not be considered in modeling: []
    13 Predictors classified...
        No variables removed since no ID or low-information variables found in data set
Since Number of Rows in data 1025 exceeds maximum, randomly sampling 1025 rows for EDA...

################ Binary_Classification problem #####################
   Columns to delete:
'   []'
   Boolean variables %s 
"   ['sex', 'fbs', 'exang']"
   Categorical variables %s 
("   ['age', 'cp', 'trestbps', 'chol', 'restecg', 'thalach', 'slope', 'ca', "
 "'thal', 'sex', 'fbs', 'exang']")
   Continuous variables %s 
"   ['oldpeak']"
   Discrete string variables %s 
'   []'
   Date and time variables %s 
'   []'
   ID variables %s 
'   []'
   Target variable %s 
'   target'
All Plots are saved in ./plots_autoviz2_png\target
Time to run AutoViz = 7 seconds
Out[11]:
age sex cp trestbps chol fbs restecg thalach exang oldpeak slope ca thal target
807 44 1 2 130 233 0 1 179 1 0.4 2 0 2 0
27 58 0 1 136 319 1 0 152 0 0.0 2 2 2 1
77 63 1 0 140 187 0 0 144 1 4.0 2 2 3 1
406 58 1 2 140 211 1 0 165 0 0.0 2 0 2 0
886 61 1 0 120 260 0 1 140 1 3.6 1 1 3 1
... ... ... ... ... ... ... ... ... ... ... ... ... ... ...
835 49 1 2 118 149 0 0 126 0 0.8 2 3 2 1
192 67 0 2 115 564 0 0 160 0 1.6 1 0 3 0
629 65 1 3 138 282 1 0 174 0 1.4 1 1 2 1
559 67 1 0 120 237 0 1 71 0 1.0 1 0 2 1
684 60 1 2 140 185 0 0 155 0 3.0 1 0 2 1

1025 rows × 14 columns

In [13]:
# 3. 지정한 폴더에 html 파일로 interactive graph 저장

AV.AutoViz(filename = '',
          dfte = df,
          depVar = 'target',
          verbose = 2, # 0: 간단히 표시, 1: 자세히 표시, 2: 파일로 저장
          max_rows_analyzed = df.shape[0],
          max_cols_analyzed = df.shape[1],
          chart_format = "html",
          save_plot_dir = "./plots_autoviz3_html")

Shape of your Data Set loaded: (1025, 14)
#######################################################################################
######################## C L A S S I F Y I N G  V A R I A B L E S  ####################
#######################################################################################
Classifying variables in data set...
Data cleaning improvement suggestions. Complete them before proceeding to ML modeling.
  Nuniques dtype Nulls Nullpercent NuniquePercent Value counts Min Data cleaning improvement suggestions
chol 152 int64 0 0.000000 14.829268 0
thalach 91 int64 0 0.000000 8.878049 0
trestbps 49 int64 0 0.000000 4.780488 0
age 41 int64 0 0.000000 4.000000 0
oldpeak 40 float64 0 0.000000 3.902439 0 skewed: cap or drop outliers
ca 5 int64 0 0.000000 0.487805 0
cp 4 int64 0 0.000000 0.390244 0
thal 4 int64 0 0.000000 0.390244 0
restecg 3 int64 0 0.000000 0.292683 0
slope 3 int64 0 0.000000 0.292683 0
sex 2 int64 0 0.000000 0.195122 0
fbs 2 int64 0 0.000000 0.195122 0
exang 2 int64 0 0.000000 0.195122 0 
  Printing upto 30 columns max in each category:
    Numeric Columns : ['oldpeak']
    Integer-Categorical Columns: ['age', 'cp', 'trestbps', 'chol', 'restecg', 'thalach', 'slope', 'ca', 'thal']
    String-Categorical Columns: []
    Factor-Categorical Columns: []
    String-Boolean Columns: []
    Numeric-Boolean Columns: ['sex', 'fbs', 'exang']
    Discrete String Columns: []
    NLP text Columns: []
    Date Time Columns: []
    ID Columns: []
    Columns that will not be considered in modeling: []
    13 Predictors classified...
        No variables removed since no ID or low-information variables found in data set
Since Number of Rows in data 1025 exceeds maximum, randomly sampling 1025 rows for EDA...

################ Binary_Classification problem #####################
   Columns to delete:
'   []'
   Boolean variables %s 
"   ['sex', 'fbs', 'exang']"
   Categorical variables %s 
("   ['age', 'cp', 'trestbps', 'chol', 'restecg', 'thalach', 'slope', 'ca', "
 "'thal', 'sex', 'fbs', 'exang']")
   Continuous variables %s 
"   ['oldpeak']"
   Discrete string variables %s 
'   []'
   Date and time variables %s 
'   []'
   ID variables %s 
'   []'
   Target variable %s 
'   target'
Saving scatterplots in HTML format
Saving distplots_cats in HTML format                                                                                   
Saving distplots_nums in HTML format                                                                                   
Saving kde_plots in HTML format                                                                                        
Saving violinplots in HTML format
Saving cat_var_plots in HTML format
Time to run AutoViz (in seconds) = 2
Out[13]:
age sex cp trestbps chol fbs restecg thalach exang oldpeak slope ca thal target
807 44 1 2 130 233 0 1 179 1 0.4 2 0 2 1
27 58 0 1 136 319 1 0 152 0 0.0 2 2 2 0
77 63 1 0 140 187 0 0 144 1 4.0 2 2 3 0
406 58 1 2 140 211 1 0 165 0 0.0 2 0 2 1
886 61 1 0 120 260 0 1 140 1 3.6 1 1 3 0
... ... ... ... ... ... ... ... ... ... ... ... ... ... ...
835 49 1 2 118 149 0 0 126 0 0.8 2 3 2 0
192 67 0 2 115 564 0 0 160 0 1.6 1 0 3 1
629 65 1 3 138 282 1 0 174 0 1.4 1 1 2 0
559 67 1 0 120 237 0 1 71 0 1.0 1 0 2 0
684 60 1 2 140 185 0 0 155 0 3.0 1 0 2 0

1025 rows × 14 columns

In [14]:
# 4. lowess 옵션 사용 jpg 저장

AV.AutoViz(filename = '',
          dfte = df,
          depVar = 'target',
          verbose = 2, # 0: 간단히 표시, 1: 자세히 표시, 2: 파일로 저장
          max_rows_analyzed = df.shape[0],
          max_cols_analyzed = df.shape[1],
          lowess = True,
          chart_format = "jpg",
          save_plot_dir = "./plots_autoviz4_jpg")

Shape of your Data Set loaded: (1025, 14)
#######################################################################################
######################## C L A S S I F Y I N G  V A R I A B L E S  ####################
#######################################################################################
Classifying variables in data set...
Data cleaning improvement suggestions. Complete them before proceeding to ML modeling.
  Nuniques dtype Nulls Nullpercent NuniquePercent Value counts Min Data cleaning improvement suggestions
chol 152 int64 0 0.000000 14.829268 0
thalach 91 int64 0 0.000000 8.878049 0
trestbps 49 int64 0 0.000000 4.780488 0
age 41 int64 0 0.000000 4.000000 0
oldpeak 40 float64 0 0.000000 3.902439 0 skewed: cap or drop outliers
ca 5 int64 0 0.000000 0.487805 0
cp 4 int64 0 0.000000 0.390244 0
thal 4 int64 0 0.000000 0.390244 0
restecg 3 int64 0 0.000000 0.292683 0
slope 3 int64 0 0.000000 0.292683 0
sex 2 int64 0 0.000000 0.195122 0
fbs 2 int64 0 0.000000 0.195122 0
exang 2 int64 0 0.000000 0.195122 0 
  Printing upto 30 columns max in each category:
    Numeric Columns : ['oldpeak']
    Integer-Categorical Columns: ['age', 'cp', 'trestbps', 'chol', 'restecg', 'thalach', 'slope', 'ca', 'thal']
    String-Categorical Columns: []
    Factor-Categorical Columns: []
    String-Boolean Columns: []
    Numeric-Boolean Columns: ['sex', 'fbs', 'exang']
    Discrete String Columns: []
    NLP text Columns: []
    Date Time Columns: []
    ID Columns: []
    Columns that will not be considered in modeling: []
    13 Predictors classified...
        No variables removed since no ID or low-information variables found in data set
Since Number of Rows in data 1025 exceeds maximum, randomly sampling 1025 rows for EDA...

################ Binary_Classification problem #####################
   Columns to delete:
'   []'
   Boolean variables %s 
"   ['sex', 'fbs', 'exang']"
   Categorical variables %s 
("   ['age', 'cp', 'trestbps', 'chol', 'restecg', 'thalach', 'slope', 'ca', "
 "'thal', 'sex', 'fbs', 'exang']")
   Continuous variables %s 
"   ['oldpeak']"
   Discrete string variables %s 
'   []'
   Date and time variables %s 
'   []'
   ID variables %s 
'   []'
   Target variable %s 
'   target'
All Plots are saved in ./plots_autoviz4_jpg\target
Time to run AutoViz = 8 seconds
Out[14]:
age sex cp trestbps chol fbs restecg thalach exang oldpeak slope ca thal target
807 44 1 2 130 233 0 1 179 1 0.4 2 0 2 0
27 58 0 1 136 319 1 0 152 0 0.0 2 2 2 1
77 63 1 0 140 187 0 0 144 1 4.0 2 2 3 1
406 58 1 2 140 211 1 0 165 0 0.0 2 0 2 0
886 61 1 0 120 260 0 1 140 1 3.6 1 1 3 1
... ... ... ... ... ... ... ... ... ... ... ... ... ... ...
835 49 1 2 118 149 0 0 126 0 0.8 2 3 2 1
192 67 0 2 115 564 0 0 160 0 1.6 1 0 3 0
629 65 1 3 138 282 1 0 174 0 1.4 1 1 2 1
559 67 1 0 120 237 0 1 71 0 1.0 1 0 2 1
684 60 1 2 140 185 0 0 155 0 3.0 1 0 2 1

1025 rows × 14 columns

In [15]:
# 5. 주피터 노트북 내 interactive graph 생성

AV.AutoViz(filename = '',
          dfte = df,
          depVar = 'target',
          verbose = 0, # 차트 포맷을 bekeh나 server, html 등으로 생성할 때는 0과 1 권장
          max_rows_analyzed = df.shape[0],
          max_cols_analyzed = df.shape[1],
          chart_format = "bokeh")

Shape of your Data Set loaded: (1025, 14)
#######################################################################################
######################## C L A S S I F Y I N G  V A R I A B L E S  ####################
#######################################################################################
Classifying variables in data set...
    Number of Numeric Columns =  1
    Number of Integer-Categorical Columns =  9
    Number of String-Categorical Columns =  0
    Number of Factor-Categorical Columns =  0
    Number of String-Boolean Columns =  0
    Number of Numeric-Boolean Columns =  3
    Number of Discrete String Columns =  0
    Number of NLP String Columns =  0
    Number of Date Time Columns =  0
    Number of ID Columns =  0
    Number of Columns to Delete =  0
    13 Predictors classified...
        No variables removed since no ID or low-information variables found in data set
Since Number of Rows in data 1025 exceeds maximum, randomly sampling 1025 rows for EDA...

################ Binary_Classification problem #####################

Time to run AutoViz (in seconds) = 1
Out[15]:
age sex cp trestbps chol fbs restecg thalach exang oldpeak slope ca thal target
807 44 1 2 130 233 0 1 179 1 0.4 2 0 2 1
27 58 0 1 136 319 1 0 152 0 0.0 2 2 2 0
77 63 1 0 140 187 0 0 144 1 4.0 2 2 3 0
406 58 1 2 140 211 1 0 165 0 0.0 2 0 2 1
886 61 1 0 120 260 0 1 140 1 3.6 1 1 3 0
... ... ... ... ... ... ... ... ... ... ... ... ... ... ...
835 49 1 2 118 149 0 0 126 0 0.8 2 3 2 0
192 67 0 2 115 564 0 0 160 0 1.6 1 0 3 1
629 65 1 3 138 282 1 0 174 0 1.4 1 1 2 0
559 67 1 0 120 237 0 1 71 0 1.0 1 0 2 0
684 60 1 2 140 185 0 0 155 0 3.0 1 0 2 0

1025 rows × 14 columns

In [16]:
# 6. 팝업창으로 interactive graph 생성

AV.AutoViz(filename = '',
          dfte = df,
          depVar = 'target',
          verbose = 0, # 차트 포맷을 bekeh나 server, html 등으로 생성할 때는 0과 1 권장
          max_rows_analyzed = df.shape[0],
          max_cols_analyzed = df.shape[1],
          chart_format = "server")

Shape of your Data Set loaded: (1025, 14)
#######################################################################################
######################## C L A S S I F Y I N G  V A R I A B L E S  ####################
#######################################################################################
Classifying variables in data set...
    Number of Numeric Columns =  1
    Number of Integer-Categorical Columns =  9
    Number of String-Categorical Columns =  0
    Number of Factor-Categorical Columns =  0
    Number of String-Boolean Columns =  0
    Number of Numeric-Boolean Columns =  3
    Number of Discrete String Columns =  0
    Number of NLP String Columns =  0
    Number of Date Time Columns =  0
    Number of ID Columns =  0
    Number of Columns to Delete =  0
    13 Predictors classified...
        No variables removed since no ID or low-information variables found in data set
Since Number of Rows in data 1025 exceeds maximum, randomly sampling 1025 rows for EDA...

################ Binary_Classification problem #####################
scatterplots can be found in URL below:
Launching server at http://localhost:54106
distplots can be found in URL below:
Launching server at http://localhost:54107
distplots can be found in URL below:
Launching server at http://localhost:54108
kde_plots can be found in URL below:
Launching server at http://localhost:54109
violinplots can be found in URL below:
Launching server at http://localhost:54110
cat_var_plots can be found in URL below:
Launching server at http://localhost:54111
Time to run AutoViz (in seconds) = 0
Out[16]:
age sex cp trestbps chol fbs restecg thalach exang oldpeak slope ca thal target
807 44 1 2 130 233 0 1 179 1 0.4 2 0 2 1
27 58 0 1 136 319 1 0 152 0 0.0 2 2 2 0
77 63 1 0 140 187 0 0 144 1 4.0 2 2 3 0
406 58 1 2 140 211 1 0 165 0 0.0 2 0 2 1
886 61 1 0 120 260 0 1 140 1 3.6 1 1 3 0
... ... ... ... ... ... ... ... ... ... ... ... ... ... ...
835 49 1 2 118 149 0 0 126 0 0.8 2 3 2 0
192 67 0 2 115 564 0 0 160 0 1.6 1 0 3 1
629 65 1 3 138 282 1 0 174 0 1.4 1 1 2 0
559 67 1 0 120 237 0 1 71 0 1.0 1 0 2 0
684 60 1 2 140 185 0 0 155 0 3.0 1 0 2 0

1025 rows × 14 columns

LazyPredict¶

https://lazypredict.readthedocs.io/en/latest/

Preparing data¶

In [17]:
y_data = df.pop('target')
x_data = df

print(x_data.shape)
print(y_data.shape)

(1025, 13)
(1025,)
In [18]:
import sklearn
print(sklearn.__version__)

1.0.2
In [19]:
from sklearn.model_selection import train_test_split

x_train, x_test, y_train, y_test = train_test_split(x_data, 
                                                    y_data, 
                                                    test_size = 0.2,
                                                    random_state = 2022,
                                                    stratify = y_data) # 클래스 비율을 동일하게 분리한다.

print(x_train.shape, y_train.shape)
print(x_test.shape, y_test.shape)

(820, 13) (820,)
(205, 13) (205,)
In [20]:
from lazypredict.Supervised import LazyClassifier

clf = LazyClassifier(verbose = 0, predictions = True)

models, predictions = clf.fit(x_train, x_test, y_train, y_test)

models

100%|██████████████████████████████████████████████████████████████████████████████████| 29/29 [00:00<00:00, 40.81it/s]
Out[20]:
Accuracy Balanced Accuracy ROC AUC F1 Score Time Taken
Model
LGBMClassifier 1.00 1.00 1.00 1.00 0.07
LabelPropagation 1.00 1.00 1.00 1.00 0.02
XGBClassifier 1.00 1.00 1.00 1.00 0.10
DecisionTreeClassifier 1.00 1.00 1.00 1.00 0.01
RandomForestClassifier 1.00 1.00 1.00 1.00 0.08
ExtraTreeClassifier 1.00 1.00 1.00 1.00 0.00
ExtraTreesClassifier 1.00 1.00 1.00 1.00 0.06
BaggingClassifier 1.00 1.00 1.00 1.00 0.02
LabelSpreading 1.00 1.00 1.00 1.00 0.03
SVC 0.94 0.94 0.94 0.94 0.02
AdaBoostClassifier 0.92 0.92 0.92 0.92 0.06
NuSVC 0.90 0.90 0.90 0.90 0.02
QuadraticDiscriminantAnalysis 0.89 0.89 0.89 0.89 0.01
GaussianNB 0.88 0.88 0.88 0.88 0.00
SGDClassifier 0.87 0.87 0.87 0.87 0.01
CalibratedClassifierCV 0.87 0.87 0.87 0.87 0.06
LogisticRegression 0.86 0.86 0.86 0.86 0.01
LinearSVC 0.86 0.86 0.86 0.86 0.03
NearestCentroid 0.86 0.86 0.86 0.86 0.01
LinearDiscriminantAnalysis 0.86 0.86 0.86 0.86 0.02
RidgeClassifier 0.86 0.86 0.86 0.86 0.01
RidgeClassifierCV 0.86 0.86 0.86 0.86 0.01
BernoulliNB 0.85 0.85 0.85 0.85 0.01
KNeighborsClassifier 0.83 0.83 0.83 0.83 0.01
PassiveAggressiveClassifier 0.78 0.78 0.78 0.78 0.01
Perceptron 0.76 0.76 0.76 0.76 0.01
DummyClassifier 0.51 0.50 0.50 0.35 0.00
In [21]:
predictions.head()
Out[21]:
AdaBoostClassifier BaggingClassifier BernoulliNB CalibratedClassifierCV DecisionTreeClassifier DummyClassifier ExtraTreeClassifier ExtraTreesClassifier GaussianNB KNeighborsClassifier ... PassiveAggressiveClassifier Perceptron QuadraticDiscriminantAnalysis RandomForestClassifier RidgeClassifier RidgeClassifierCV SGDClassifier SVC XGBClassifier LGBMClassifier
0 0 0 0 0 0 1 0 0 0 0 ... 0 0 0 0 0 0 0 0 0 0
1 1 1 0 1 1 1 1 1 1 0 ... 1 1 1 1 1 1 1 1 1 1
2 1 1 1 1 1 1 1 1 1 1 ... 1 1 1 1 1 1 1 1 1 1
3 1 1 1 1 1 1 1 1 1 1 ... 1 1 1 1 1 1 1 1 1 1
4 1 0 1 1 0 1 0 0 1 1 ... 1 1 0 0 1 1 1 0 0 0

5 rows × 27 columns

In [22]:
from sklearn.metrics import classification_report

for model_name in predictions.columns.tolist():
    print(f'{model_name}')
    print(classification_report(y_test, predictions[model_name]))

AdaBoostClassifier
              precision    recall  f1-score   support

           0       0.95      0.89      0.92       100
           1       0.90      0.95      0.93       105

    accuracy                           0.92       205
   macro avg       0.92      0.92      0.92       205
weighted avg       0.92      0.92      0.92       205

BaggingClassifier
              precision    recall  f1-score   support

           0       1.00      1.00      1.00       100
           1       1.00      1.00      1.00       105

    accuracy                           1.00       205
   macro avg       1.00      1.00      1.00       205
weighted avg       1.00      1.00      1.00       205

BernoulliNB
              precision    recall  f1-score   support

           0       0.88      0.80      0.84       100
           1       0.82      0.90      0.86       105

    accuracy                           0.85       205
   macro avg       0.85      0.85      0.85       205
weighted avg       0.85      0.85      0.85       205

CalibratedClassifierCV
              precision    recall  f1-score   support

           0       0.95      0.78      0.86       100
           1       0.82      0.96      0.89       105

    accuracy                           0.87       205
   macro avg       0.89      0.87      0.87       205
weighted avg       0.88      0.87      0.87       205

DecisionTreeClassifier
              precision    recall  f1-score   support

           0       1.00      1.00      1.00       100
           1       1.00      1.00      1.00       105

    accuracy                           1.00       205
   macro avg       1.00      1.00      1.00       205
weighted avg       1.00      1.00      1.00       205

DummyClassifier
              precision    recall  f1-score   support

           0       0.00      0.00      0.00       100
           1       0.51      1.00      0.68       105

    accuracy                           0.51       205
   macro avg       0.26      0.50      0.34       205
weighted avg       0.26      0.51      0.35       205

ExtraTreeClassifier
              precision    recall  f1-score   support

           0       1.00      1.00      1.00       100
           1       1.00      1.00      1.00       105

    accuracy                           1.00       205
   macro avg       1.00      1.00      1.00       205
weighted avg       1.00      1.00      1.00       205

ExtraTreesClassifier
              precision    recall  f1-score   support

           0       1.00      1.00      1.00       100
           1       1.00      1.00      1.00       105

    accuracy                           1.00       205
   macro avg       1.00      1.00      1.00       205
weighted avg       1.00      1.00      1.00       205

GaussianNB
              precision    recall  f1-score   support

           0       0.89      0.85      0.87       100
           1       0.86      0.90      0.88       105

    accuracy                           0.88       205
   macro avg       0.88      0.88      0.88       205
weighted avg       0.88      0.88      0.88       205

KNeighborsClassifier
              precision    recall  f1-score   support

           0       0.85      0.80      0.82       100
           1       0.82      0.87      0.84       105

    accuracy                           0.83       205
   macro avg       0.84      0.83      0.83       205
weighted avg       0.84      0.83      0.83       205

LabelPropagation
              precision    recall  f1-score   support

           0       1.00      1.00      1.00       100
           1       1.00      1.00      1.00       105

    accuracy                           1.00       205
   macro avg       1.00      1.00      1.00       205
weighted avg       1.00      1.00      1.00       205

LabelSpreading
              precision    recall  f1-score   support

           0       1.00      1.00      1.00       100
           1       1.00      1.00      1.00       105

    accuracy                           1.00       205
   macro avg       1.00      1.00      1.00       205
weighted avg       1.00      1.00      1.00       205

LinearDiscriminantAnalysis
              precision    recall  f1-score   support

           0       0.96      0.74      0.84       100
           1       0.80      0.97      0.88       105

    accuracy                           0.86       205
   macro avg       0.88      0.86      0.86       205
weighted avg       0.88      0.86      0.86       205

LinearSVC
              precision    recall  f1-score   support

           0       0.95      0.76      0.84       100
           1       0.81      0.96      0.88       105

    accuracy                           0.86       205
   macro avg       0.88      0.86      0.86       205
weighted avg       0.88      0.86      0.86       205

LogisticRegression
              precision    recall  f1-score   support

           0       0.93      0.78      0.85       100
           1       0.82      0.94      0.88       105

    accuracy                           0.86       205
   macro avg       0.87      0.86      0.86       205
weighted avg       0.87      0.86      0.86       205

NearestCentroid
              precision    recall  f1-score   support

           0       0.94      0.76      0.84       100
           1       0.81      0.95      0.87       105

    accuracy                           0.86       205
   macro avg       0.87      0.86      0.86       205
weighted avg       0.87      0.86      0.86       205

NuSVC
              precision    recall  f1-score   support

           0       0.94      0.85      0.89       100
           1       0.87      0.95      0.91       105

    accuracy                           0.90       205
   macro avg       0.91      0.90      0.90       205
weighted avg       0.91      0.90      0.90       205

PassiveAggressiveClassifier
              precision    recall  f1-score   support

           0       0.77      0.77      0.77       100
           1       0.78      0.78      0.78       105

    accuracy                           0.78       205
   macro avg       0.78      0.78      0.78       205
weighted avg       0.78      0.78      0.78       205

Perceptron
              precision    recall  f1-score   support

           0       0.76      0.74      0.75       100
           1       0.76      0.77      0.76       105

    accuracy                           0.76       205
   macro avg       0.76      0.76      0.76       205
weighted avg       0.76      0.76      0.76       205

QuadraticDiscriminantAnalysis
              precision    recall  f1-score   support

           0       0.91      0.87      0.89       100
           1       0.88      0.91      0.90       105

    accuracy                           0.89       205
   macro avg       0.89      0.89      0.89       205
weighted avg       0.89      0.89      0.89       205

RandomForestClassifier
              precision    recall  f1-score   support

           0       1.00      1.00      1.00       100
           1       1.00      1.00      1.00       105

    accuracy                           1.00       205
   macro avg       1.00      1.00      1.00       205
weighted avg       1.00      1.00      1.00       205

RidgeClassifier
              precision    recall  f1-score   support

           0       0.96      0.74      0.84       100
           1       0.80      0.97      0.88       105

    accuracy                           0.86       205
   macro avg       0.88      0.86      0.86       205
weighted avg       0.88      0.86      0.86       205

RidgeClassifierCV
              precision    recall  f1-score   support

           0       0.96      0.74      0.84       100
           1       0.80      0.97      0.88       105

    accuracy                           0.86       205
   macro avg       0.88      0.86      0.86       205
weighted avg       0.88      0.86      0.86       205

SGDClassifier
              precision    recall  f1-score   support

           0       0.91      0.82      0.86       100
           1       0.84      0.92      0.88       105

    accuracy                           0.87       205
   macro avg       0.88      0.87      0.87       205
weighted avg       0.88      0.87      0.87       205

SVC
              precision    recall  f1-score   support

           0       0.97      0.90      0.93       100
           1       0.91      0.97      0.94       105

    accuracy                           0.94       205
   macro avg       0.94      0.94      0.94       205
weighted avg       0.94      0.94      0.94       205

XGBClassifier
              precision    recall  f1-score   support

           0       1.00      1.00      1.00       100
           1       1.00      1.00      1.00       105

    accuracy                           1.00       205
   macro avg       1.00      1.00      1.00       205
weighted avg       1.00      1.00      1.00       205

LGBMClassifier
              precision    recall  f1-score   support

           0       1.00      1.00      1.00       100
           1       1.00      1.00      1.00       105

    accuracy                           1.00       205
   macro avg       1.00      1.00      1.00       205
weighted avg       1.00      1.00      1.00       205
In [23]:
from lightgbm import LGBMClassifier
from sklearn.metrics import accuracy_score

lgbm = LGBMClassifier()
lgbm.fit(x_train, y_train)
y_pred = lgbm.predict(x_test)
accuracy_score(y_pred, y_test)
Out[23]:
1.0

Scikit-optimize¶

파라미터 자동 튜닝

SVM 튜닝없이 돌렸을 때

In [24]:
import numpy as np
from sklearn.svm import SVC
from sklearn.pipeline import Pipeline
from sklearn.preprocessing import StandardScaler

X_train = pd.DataFrame(x_train)
X_test = pd.DataFrame(x_test)

pipe = Pipeline(steps = [("preprocessor", StandardScaler()), ("classifier", SVC())])

pipe.fit(x_train, y_train)

y_pred = pipe.predict(x_test)

accuracy_score(y_pred, y_test)
Out[24]:
0.9365853658536586
In [25]:
pipe['classifier'].get_params()
Out[25]:
{'C': 1.0,
 'break_ties': False,
 'cache_size': 200,
 'class_weight': None,
 'coef0': 0.0,
 'decision_function_shape': 'ovr',
 'degree': 3,
 'gamma': 'scale',
 'kernel': 'rbf',
 'max_iter': -1,
 'probability': False,
 'random_state': None,
 'shrinking': True,
 'tol': 0.001,
 'verbose': False}

BayesSearchCV 이용 파라미터 튜닝¶

In [26]:
from skopt import BayesSearchCV
In [27]:
pipe = Pipeline(steps = [("preprocessor", StandardScaler()), ("classifier", SVC())])

opt = BayesSearchCV(
        pipe, 
        {
            'classifier__C': (1e-1, 1e+1, 'log-uniform'),
            'classifier__gamma': (1e-6, 1e+1, 'log-uniform'),
            'classifier__degree': (1, 8), # integer valued parameter
            'classifier__kernel': ['linear', 'poly', 'rbf'], # categorical parameter
        },
        n_iter = 8,
        cv = 3
)

opt.fit(x_train, y_train)

print("val. score: %s" % opt.best_score_)
print("test score: %s " % opt.score(x_test, y_test))

val. score: 0.9670686167036533
test score: 1.0
In [28]:
opt.best_params_
Out[28]:
OrderedDict([('classifier__C', 0.3190547286265487),
             ('classifier__degree', 4),
             ('classifier__gamma', 0.632736824293571),
             ('classifier__kernel', 'poly')])

Lazy Predict Usage¶

In [29]:
import lazypredict

Classification¶

In [30]:
from lazypredict.Supervised import LazyClassifier
from sklearn.datasets import load_breast_cancer
from sklearn.model_selection import train_test_split

data = load_breast_cancer()

# X data, Y data split
X = data.data
y = data.target

X_train, X_test, y_train, y_test = train_test_split(X, y, test_size = .5, random_state = 123)
clf = LazyClassifier(verbose = 0, ignore_warnings = True, custom_metric = None)
models, predictions = clf.fit(X_train, X_test, y_train, y_test)
models

100%|██████████████████████████████████████████████████████████████████████████████████| 29/29 [00:00<00:00, 57.18it/s]
Out[30]:
Accuracy Balanced Accuracy ROC AUC F1 Score Time Taken
Model
LinearSVC 0.99 0.99 0.99 0.99 0.01
Perceptron 0.99 0.98 0.98 0.99 0.00
LogisticRegression 0.99 0.98 0.98 0.99 0.01
SVC 0.98 0.98 0.98 0.98 0.01
XGBClassifier 0.98 0.98 0.98 0.98 0.07
LabelPropagation 0.98 0.97 0.97 0.98 0.01
LabelSpreading 0.98 0.97 0.97 0.98 0.01
BaggingClassifier 0.97 0.97 0.97 0.97 0.02
PassiveAggressiveClassifier 0.98 0.97 0.97 0.98 0.00
SGDClassifier 0.98 0.97 0.97 0.98 0.01
RandomForestClassifier 0.97 0.97 0.97 0.97 0.08
CalibratedClassifierCV 0.98 0.97 0.97 0.98 0.02
LGBMClassifier 0.97 0.97 0.97 0.97 0.05
QuadraticDiscriminantAnalysis 0.96 0.97 0.97 0.97 0.01
ExtraTreesClassifier 0.97 0.96 0.96 0.97 0.06
RidgeClassifierCV 0.97 0.96 0.96 0.97 0.01
RidgeClassifier 0.97 0.96 0.96 0.97 0.01
AdaBoostClassifier 0.96 0.96 0.96 0.96 0.06
KNeighborsClassifier 0.96 0.96 0.96 0.96 0.01
BernoulliNB 0.95 0.95 0.95 0.95 0.01
LinearDiscriminantAnalysis 0.96 0.95 0.95 0.96 0.01
GaussianNB 0.95 0.95 0.95 0.95 0.00
NuSVC 0.95 0.94 0.94 0.95 0.01
ExtraTreeClassifier 0.94 0.93 0.93 0.94 0.01
NearestCentroid 0.95 0.93 0.93 0.95 0.01
DecisionTreeClassifier 0.93 0.93 0.93 0.93 0.01
DummyClassifier 0.64 0.50 0.50 0.50 0.00

Regression¶

In [31]:
from lazypredict.Supervised import LazyRegressor
from sklearn import datasets
from sklearn.utils import shuffle
import numpy as np

boston = datasets.load_boston()
X, y = shuffle(boston.data, boston.target, random_state = 13)
X = X.astype(np.float32)
In [32]:
offset = int(X.shape[0] * 0.9)
X_train, y_train = X[:offset], y[:offset]
X_test, y_test = X[offset:], y[offset:]
reg = LazyRegressor(verbose = 0, ignore_warnings = False, custom_metric = None)
models2, predictions2 = reg.fit(X_train, X_test, y_train, y_test)

 14%|███████████▊                                                                       | 6/42 [00:00<00:00, 56.45it/s]
ElasticNetCV model failed to execute
Gram matrix passed in via 'precompute' parameter did not pass validation when a single element was checked - please check that it was computed properly. For element (6,7) we computed -263.0660400390625 but the user-supplied value was -263.06610107421875.

 60%|████████████████████████████████████████████████▊                                 | 25/42 [00:00<00:00, 29.42it/s]
LassoCV model failed to execute
Gram matrix passed in via 'precompute' parameter did not pass validation when a single element was checked - please check that it was computed properly. For element (6,7) we computed -263.0660400390625 but the user-supplied value was -263.06610107421875.

100%|██████████████████████████████████████████████████████████████████████████████████| 42/42 [00:03<00:00, 12.82it/s]
In [33]:
models2
Out[33]:
Adjusted R-Squared R-Squared RMSE Time Taken
Model
SVR 0.83 0.88 2.62 0.01
BaggingRegressor 0.83 0.88 2.63 0.02
NuSVR 0.82 0.86 2.76 0.01
RandomForestRegressor 0.81 0.86 2.78 0.19
XGBRegressor 0.81 0.86 2.79 0.08
GradientBoostingRegressor 0.81 0.86 2.84 0.08
ExtraTreesRegressor 0.79 0.84 2.98 0.13
AdaBoostRegressor 0.78 0.83 3.04 0.06
HistGradientBoostingRegressor 0.77 0.83 3.06 0.39
PoissonRegressor 0.77 0.83 3.11 0.01
LGBMRegressor 0.77 0.83 3.11 0.06
KNeighborsRegressor 0.77 0.83 3.12 0.01
DecisionTreeRegressor 0.65 0.74 3.79 0.01
MLPRegressor 0.65 0.74 3.80 0.35
HuberRegressor 0.64 0.74 3.84 0.02
GammaRegressor 0.64 0.73 3.88 0.01
LinearSVR 0.62 0.72 3.96 0.01
RidgeCV 0.62 0.72 3.97 0.01
BayesianRidge 0.62 0.72 3.97 0.00
Ridge 0.62 0.72 3.97 0.01
LinearRegression 0.62 0.72 3.97 0.00
TransformedTargetRegressor 0.62 0.72 3.97 0.00
LassoLarsIC 0.62 0.72 3.98 0.01
LassoLarsCV 0.62 0.72 3.98 0.01
Lars 0.61 0.72 3.99 0.01
LarsCV 0.61 0.71 4.02 0.03
SGDRegressor 0.60 0.70 4.07 0.01
TweedieRegressor 0.59 0.70 4.12 0.00
ElasticNet 0.58 0.69 4.16 0.01
Lasso 0.54 0.66 4.35 0.01
RANSACRegressor 0.53 0.65 4.41 0.04
OrthogonalMatchingPursuitCV 0.45 0.59 4.78 0.01
PassiveAggressiveRegressor 0.37 0.54 5.09 0.00
GaussianProcessRegressor 0.23 0.43 5.65 0.02
OrthogonalMatchingPursuit 0.16 0.38 5.89 0.00
ExtraTreeRegressor 0.08 0.32 6.17 0.01
QuantileRegressor -0.35 -0.00 7.49 1.60
DummyRegressor -0.38 -0.02 7.56 0.00
LassoLars -0.38 -0.02 7.56 0.00
KernelRidge -11.50 -8.25 22.74 0.02